Poems and Limericks Still Break AI Safety Guardrails

A researcher claims to have jailbroken OpenAI, Google, and Anthropic models using nothing more than creative writing prompts — suggesting safety alignment remains brittle against low-sophistication attacks.

@BrianRoemmele posted that he successfully bypassed safety filters at OpenAI, Google, and Anthropic by framing restricted requests as poems and limericks. The post gathered 469 likes and a lively thread. While jailbreak claims require independent verification, the general finding aligns with a well-documented pattern: safety guardrails trained against direct requests often fail against creative reframing.

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum. Vestibulum ante ipsum primis in faucibus orci luctus et ultrices posuere cubilia curae.

Unlock the full briefing

Get every story in today's briefing, the full archive, and the daily AI intelligence brief.

All stories today

Full archive

Daily brief